RAM Advanse Analysis Error - "Solution cannot converge"

[imsengr]

Hi all,
I am running my second model in RAM Advanse, a tall, wide but slim structural frame with dead (gravity) and lateral loads from seismic and wind. When I run analysis, I am getting this error, "The solution cannot be converged during iterations. Analysis terminated." What does the error mean? What does it mean when it says it cannot converge? Is there something or some changes that I need to make. Any and all help welcomed. Thanks.

 

[UcfSE]

Your frame is probably too flexible, or you have a mistake in your model, such as members not properly connected to nodes or supports. An analysis that does not converge is one that results in larger and larger changes in loads or deformations with each successive iteration instead of those getting smaller and smaller.

Remember to check your help file also. It may actually help.

 

[JAE]

I would suggest you simply call RAM's technical staff and work with them. In most all cases, that is the best way to get it resolved.

If an analysis model is interating and trying to converge, what that tells me is that it is running a form of P-delta analysis and the model is not properly braced or configured....usually too many pins, not enough bracing, or incorrect boundary conditions.

Call them - they've always been very helpful with us when we run into a problem.

 

[Tomfh]

It means your structure fell over.

 

[imsengr]

Thanks for all your suggestions and replies, esp. UcfSE and JAE. Please keep them coming. Tomfh, I am concerned that you think my structure fell over. I certainly do not want that to happen, and I would be grateful if you share some thoughts and advice on how I can avoid that.

Per all your advice, I did try RAM Technical Support and here's what I found out. You're right, they're really helpful and the guy actually called me at the office!

I had modeled all my diagonal bracing as Tension only members, and ran my analysis in the p-delta mode (second order analysis). If I had compressive forces in my cross bracing (which is certainly possible), RAM could not use those diagonal bracing. It will be as if those cross bracings are not even there! RAM would then find it very difficult to converge while seeking a solution, and that was it generated that error, "The Solution cannot be converged during iterations. Analysis terminated."

Convergence is apparently an equilibrium state that is achieved if the model ran properly. Could someone elaborate on "convergence", please?

Another possibility is running it in p-delta mode. BTW, can someone please elaborate more on what p-delta mode (second order analysis) is and how it diffes from the other mode (first analysis)? Anyway, when running it in p-delta mode, RAM would not be able to converge if there were large node deflections in my structure (certainly a possibility in my preliminary frame).

Anyway, RAM Technical Support advised me to try and run my structure without using the Tension Only members for my diagonal bracing, and also to uncheck the p-delta mode analysis. I was also told to check the box that reads, "Cancel analysis if any of the load cases fail to converge".

If my model ran properly with the "convergence" errors, then it is possible, that Tension Only Members and p-delta analysis is causing the error. It could be that my model is perfectly fine, but turning those two attributes is causing the error. My model could also have flaws that I need to investigate and correct.

I could then run my model again initially turning on Tension Members only to see if it works, and if it does, to run p-delta mode.

Does that ring a bell in anyone? If so, and if you can elaborate on what RAM Technical Support advised, please share your thoughts and advice, ok?

Thanks, y'all and have a great day.

 

[UcfSE]

"Convergence" and "P-DELTA" and "P-delta" are some fairly basic terms when it comes to analysis of structures. You should have a good understanding of those concepts before you try to model with software. As they say, garbage in equals garbage out. Your best bet is to find some good structural analysis books including matrix analysis and finite elements. Texts will have the information you need. Try Kassmimali for matrix analysis and Robert Cook, et. al. for FEA, for starters.